Purchase this article with an account.
Juleke Eugenie Anne Majoor, Koenraad Arndt Vermeer, Hans G Lemij; Contrast-to-noise ratios to evaluate the detection of progression in eyes with diffuse and local glaucomatous damage. Invest. Ophthalmol. Vis. Sci. 2019;60(9):6154.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
It is difficult to determine which structural or functional parameter is best for detecting progression in a general glaucoma patient in a certain stage of the disease. Our aim was to test whether the sensitivity for detecting disease progression differs between eyes with diffuse and local glaucomatous damage but with the similar average visual field loss. We used contrast-to-noise ratios (CNRs) to assess any such differences.
CNRs for the retinal nerve fiber layer thickness (RNFLt) from optical coherence tomography (OCT) and mean deviation (MD) from standard automated perimetry (SAP) were determined in longitudinal data from glaucomatous eyes from a prospective study (205 eyes, 123 participants). Contrast was defined as the change in the parameter between two consecutive glaucoma stages, and noise was the measurement variability of the parameter in these two stages. The higher the CNR, the more sensitive the parameter is to detect progression from one stage to the next. Eight different subgroups of local and diffuse damage were defined based on eight different offsets (0 dB to ±1.75 dB) from a LOESS fit on the MD (Figure 1). CNRs for each parameter and each subgroup were calculated for the transition of mild to moderate and moderate to advanced glaucoma.
The CNRs have been shown in Figure 2. Overall, we noticed an increase in the CNRs for both the MD and RNFLt as the damage became either more diffuse (up to ‘LOESS – 1.75 dB’) or more local (up to ‘LOESS + 1.75 dB’). The CNR for RNFLt in locally damaged eyes was generally higher than the CNR for MD. By contrast, the CNR in diffusely damaged eyes was generally higher for MD than for RNFLt. This could be explained for RNFLt by the less pronounced floor effect in eyes with local damage where the healthy parts have not yet reached the limit (contrast ↑), and for MD by the instruments limit of <0dB in eyes with diffuse damage where more thresholds have reached this limit (noise ↓).
The sensitivity of structural and functional measurements for detecting disease progression differs between eyes with diffuse and those with local glaucomatous damage. Furthermore, the results indicate that for patients with similar MDs, OCT is more sensitive for detecting progression in eyes with local glaucomatous damage than SAP, while SAP is more sensitive for detecting progression in eyes with diffuse damage.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only